In practice, calls to main are not meaningful:
- If main is called in live code, it's recursive
- If main is called in dead code, it's eliminated
However, if optimization/dead-stripping is turned off, the dead-code
case emits bad code in our GLSL/Metal backends. Rather than add a
special case to the backends to work around an error in a meaningless
edge case, we now reject the function call entirely.
Prototyping main() is, as far as I can tell, harmless, so this
continues to be allowed.
Change-Id: I69916840c06810cd948bd74ff168ada1c4fc74f8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/557578
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Arman Uguray <armansito@google.com>
Commit-Queue: Arman Uguray <armansito@google.com>
SkGlyphDigest encompasses glyph attributes used early in the
text drawing stack. This is a newer class, so move functions and
constants to SkGlyphDigest that were historically misplaced.
Change-Id: Iec20c9a96fd7f4adf560f0c385dfd90abf13f009
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/557579
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Herb Derby <herb@google.com>
Bug: skia:13438
Change-Id: Id1b7e6759144820f200040616deb3a120d809765
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554405
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
The fuzzer has noticed that we didn't enforce this. Now we issue an
error if directives don't have a newline at the end.
Change-Id: I29064bec2a2be628d87b7286cd16ac0089713413
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/557380
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Leaving this private for now, so we can address all of the testing
issues. Once we decide on the best ergonomics, this can be moved to
a public API.
Bug: skia:11209
Change-Id: I030e223d8cbd16438d9dfd518af41e1a7269b6dc
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/550701
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
Extract the functionality of using either SkStrikeForGPU*
or sk_sp<SkStrike> for handling descriptors from either RemoteStrike
or from SkStrike. This will be used by the PathOpSubmitter and
DrawableOpSubmitter in the future.
Change-Id: I672ac9439a62c0485ac44b0a676640c30042e076
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/557056
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Change-Id: I16544e81c170b38d97a1f7716e4fc5bb4aa05721
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556607
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
GrBackendTexture accepts an optional label param. Currently it was
passed only to Gl backend. This CL extends the scope and pass the
optional param, that comes from external clients, to other backends
like Dawn, Mtl, D3D etc.
Bug: chromium:1164111
Change-Id: Ibcaa3f7b3a1e294c5ab68e2a621df43c8c714d7a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556757
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
This has been replaced by findOrCreateRuntimeEffectSnippet.
Change-Id: I26e1f85c74071b14b75ab86410e4cc30a81bf873
Bug: skia:13405
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/552716
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Large startAngle values (combined w/ an almost 360 sweep angle) can
cause an infinite loop in angles_to_unit_vectors in the loop that
tweaks the stop vector.
Bug: oss-fuzz:19916
Bug: oss-fuzz:48241
Change-Id: I914aac45aeeb302f23dc5b4b669182fd985cba4f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556696
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Added code to remove code within switch statements due to break, return,
and continue statements. The logic is applied conservatively and only
among the statements of an individual switch-case statement without
affecting other cases.
Bug: skia:13484
Change-Id: Id5b936ca91e562a5180a31a039a85de9e093961d
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556376
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Arman Uguray <armansito@google.com>
The FP factory would apply optimizations based on the mode, and return
one of the passed-in FPs. We use 'nullptr' as the 'src' FP when
constructing the blender's FP in make_effect_fp. This means that we get
back 'nullptr' from a src-mode blender. Later, we interpret that as
src-over (the default for an unset blender). Oops.
The new test variant would previously fail, before the fix to the FP.
The tweak to the FP technically eliminates an optimization, but it's one
that only applies to blending happening in the shader (eg, compose
shader, or runtime effects using a blender), and only when the blender
is one of the trivial modes. The resulting shader will still optimize
down, it just involves a bit of extra work before that happens. This
shouldn't have any long-term performance impact, particularly on
important scenarios.
Change-Id: Id5c6a6ca8a263b35c2dca3c41171748cffd41adb
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556599
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Pass an optional label parameter to GrBackendTexture and pass
it to MakeWrapped of GrGLTexture because almost all textures
that are coming clients are things where the client owns the
gl texture and passes it into Skia as a wrapped texture via a
GrBackendTexture.
Bug: chromium:1164111
Change-Id: I4bfddda956c72b53d0070595ef3268ee1a2b747f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/555597
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
This checks the padding of a matrix followed by a scalar/vector, and the
padding of a scalar/vector followed by a matrix. The checks support
all mixes of 16- and 32-bit sized elements.
After some investigation, this CL also removes some TODOs in
`get_ubo_aligned_offset`. Our uniform system does not support structs,
which is a large source of disparity between layouts. With structs
removed from the equation, the only difference between layouts seems to
be related to std140 array padding. (std430 and Metal seem to be
entirely the same.)
Change-Id: I76c48ae1e597b98aad8a8f8495ab4a8ad262845b
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556356
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
The fuzzer figured out that #version 300 would let you declare a
nonsquare matrix uniform. Quite a bit of downstream code isn't ready for
that, yet. For now, just tighten things up so the var declaration checks
match the types supported by SkRuntimeEffect::Uniform.
Bug: oss-fuzz:48829
Change-Id: I63daf3dfa7deb795901f19553805cf2351378620
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556359
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
We have unit tests that require api 26 to run, but we have no vulkan
bots that would run them. API 26 is supported on all O devices and I
don't think we'll ever want to run Vulkan on previous android devices.
So I think it is fine for us to just test at api 26 for all our test
bots.
Change-Id: I8f92af6504960b7b688281ad71f5f307fdf57f49
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556028
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
Thankfully, this was already working; no changes in the uniform manager
were needed.
Change-Id: Ic2c4807e8efa63a05127d6f96d8a58ce785bbc1e
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556316
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Previously, we would treat double values outside [-FLT_MAX, FLT_MAX]
as finite. In practice, this introduces many hazards; any place in the
code which handled the double value as a float would silently convert
the value to infinity. This includes high-traffic calls like
Literal::MakeFloat.
Note that the if checks are structured in a slightly awkward way to
ensure that NaNs are treated as non-finite.
The original buggy behavior can be seen at http://review.skia.org/556078
Change-Id: Ic126afe57c3d6c7aa3edf9c8f7e339abc5f77739
Bug: oss-fuzz:48592
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/556080
Reviewed-by: Arman Uguray <armansito@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
The fuzzer managed to create a NaN using a carefully-crafted mix of
intrinsics and constant folding. (`cosh(421)` is a very large double,
which becomes +Inf when cast to float, which is then multiplied by 0;
zero times infinity is NaN.)
Our code which checked to see if a value is in range of an int did not
consider NaNs and their always-false behavior, so it incorrectly
decided that NaN was in range. This CL reverses the check so that a NaN
will not pass, but all other values will behave the same.
Followup CLs should probably also tighten up the folding/optimizer
behavior so that NaNs/Infs are not created at all.
Change-Id: Idd2b0447ebe115e00bdba63ca7ff655f6c902fc6
Bug: oss-fuzz:48592
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/555009
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
The UniformManager would previously assert if it detected an alignment
error, but did not actually have any mechanism for automatically adding
padding between misaligned elements. If a user specifies a uniform
layout like `uniform float a; uniform float4 b;`, we now insert padding
between `a` and `b` so that `b` will be properly aligned. (This reuses
the logic that was originally used to trigger the alignment assertion.)
This CL fixes the bug and adds a test. The test is only active for
elements of matching size; mixed half- and full-precision elements
will need improved test logic.
Change-Id: I8c04eb6350fa73bdbcd1a08e1a45b17fee0d4194
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/555440
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
This is a reland of commit a9b52ca52a
Original change's description:
> Add option to GrGpuBuffer::updateData to *not* discard non-updated area.
>
> Also support minimal map region in Dawn implementation when onUpdateBuffer uses mapping.
>
> Bug: skia:13427
> Change-Id: I5c8a2872b520e04f8a55085c86430e635da2f43f
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553585
> Reviewed-by: Greg Daniel <egdaniel@google.com>
> Commit-Queue: Brian Salomon <bsalomon@google.com>
Bug: skia:13427
Change-Id: I04e8a4cba5dd3d77e457a4df2d8c20397c2ae760
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554997
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
The Stride() function in the UniformManager, when given a matrix type,
was intended to return the stride of a single matrix column. However, it
accidentally called `Stride(1)` instead of `Stride(kNonArray)`, which
meant that it was returning the stride for a `vec2[1]` (four floats)
instead of the stride of a `vec2` (two floats). Interestingly, this
still returned the correct answer for mat3 and mat4, because `vec3`,
`vec4`, `vec3[1]` and `vec4[1]` all have the same stride--four floats.
This CL fixes the bug and adds a test.
Change-Id: I9c06b7da7253a86c1d9545c7e177bc916b49c9b9
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/555161
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
std140, std430, and Metal all agree that non-array vec3 types should
pack like the equivalent vec4.
std140: http://screen/5NHiYze3sk4CbWv
std430: http://screen/8G6CuomyEkxge5t
Metal: http://screen/7t9LfvmuZmtJ4zq
Our Stride() function in the UniformManager, however, did not pad out
vec3 types. This would manifest as an assertion when the _next_
uniform was added; we would assert because get_ubo_aligned_offset
would mismatch our expected values.
This CL fixes the bug and adds a test.
Change-Id: I33f04f37d68b0099236576f69250ed73d9e010cd
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554404
Commit-Queue: John Stiles <johnstiles@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
The pattern of treating Metal half-precision uniforms differently was
affecting both existing and upcoming tests, so it's factored out to a
helper function.
Also, added in calls to `doneWithExpectedUniforms`--these aren't
required, but it will give us more thorough test coverage.
Change-Id: I3f76a775816242cd1e5de9be1a96486738ddb12f
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554998
Reviewed-by: James Godfrey-Kittle <jamesgk@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: James Godfrey-Kittle <jamesgk@google.com>
Similar to floats, Metal expects short-int uniforms to be passed in 16
bits, but other layouts always use 32 bits for integers.
Change-Id: I99575349d8547876ab6dab00f393158fbeea5385
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554345
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Bug: skia:13430
Change-Id: I4958b0d14f7805889ed2061fe929778f69d40c48
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/549566
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
Also support minimal map region in Dawn implementation when onUpdateBuffer uses mapping.
Bug: skia:13427
Change-Id: I5c8a2872b520e04f8a55085c86430e635da2f43f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553585
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
This test verifies that float scalars and vectors can be added as
uniforms. In Metal, this is a particularly interesting challenge because
we represent half values as 16-bit floats.
Change-Id: I4ce2280a61afee905ac5980792b6dbe6e4be572f
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554344
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
This reverts commit 6c05f9064f.
Reason for revert: fixed in release
Original change's description:
> Revert "Create simple unit test for UniformManager types and layouts."
>
> This reverts commit c42782ad74.
>
> Reason for revert: breaks in release
>
> Original change's description:
> > Create simple unit test for UniformManager types and layouts.
> >
> > The first test is an extremely simple check (can we create a uniform of
> > every allowed type in every layout?) and already uncovered a bug, so
> > we're off to a good start.
> >
> > Change-Id: If9d6aa9c0845727c5422185094711734951c2e0a
> > Bug: skia:13478
> > Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554339
> > Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> > Auto-Submit: John Stiles <johnstiles@google.com>
> > Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> > Commit-Queue: John Stiles <johnstiles@google.com>
>
> Bug: skia:13478
> Change-Id: I75c051c3daee63cad0ffd55e3823d9eb333980ee
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554398
> Auto-Submit: John Stiles <johnstiles@google.com>
> Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bug: skia:13478
Change-Id: Ic8fafda7c2149cbb07e3469fe7165787e110114e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554400
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
This reverts commit c42782ad74.
Reason for revert: breaks in release
Original change's description:
> Create simple unit test for UniformManager types and layouts.
>
> The first test is an extremely simple check (can we create a uniform of
> every allowed type in every layout?) and already uncovered a bug, so
> we're off to a good start.
>
> Change-Id: If9d6aa9c0845727c5422185094711734951c2e0a
> Bug: skia:13478
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554339
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> Auto-Submit: John Stiles <johnstiles@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> Commit-Queue: John Stiles <johnstiles@google.com>
Bug: skia:13478
Change-Id: I75c051c3daee63cad0ffd55e3823d9eb333980ee
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554398
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
This will make our Bazel rules in G3 easier to use.
We will have to clean up a lot of clients (and this still
might require a manual G3 roll to land to account for the
source files).
Eventually, we will be able to delete the old one
//include/third_party/skcms and //third_party/skcms
Bug: skia:12451, b/237076898
Change-Id: I9fd55607cbb7e1196d175aa8f140e99a73505c89
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554056
Reviewed-by: Brian Salomon <bsalomon@google.com>
The first test is an extremely simple check (can we create a uniform of
every allowed type in every layout?) and already uncovered a bug, so
we're off to a good start.
Change-Id: If9d6aa9c0845727c5422185094711734951c2e0a
Bug: skia:13478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/554339
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Bug: skia:13357
Change-Id: Ib2566e9c38f667368b20f18ed2b1851615602cad
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/548480
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
Bug: skia:13427
Change-Id: I05e538bb9464d32817a39c8d38e53c2b284c2960
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553357
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Bug: skia:13430
Change-Id: I6c577b459bc3a699e6c660d21d2b119643fe1f88
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553877
Reviewed-by: Eric Boren <borenet@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Turns out sized arrays don't work currently; the path of least
resistance is simply disabling them for now. See bug for explanation.
Bug: skia:13471
Change-Id: I90e321ccae199ac5697d1d9a63f73e30554e2981
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553583
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Reviewed-by: Arman Uguray <armansito@google.com>
Graphite expects uniforms to be passed in as a span of SkUniforms.
When making a runtime effect, we now convert the SkRuntimeEffect's
`Uniform` data into SkUniforms as well.
I briefly looked at sharing a single uniform type, but
SkRuntimeEffect::Uniform is public API, so changing it is non-trivial,
and Uniform would not be usable as a direct replacement for SkUniform.
(Uniform is non-POD since it contains an SkString; SkUniform is POD
and we declare a bunch of them at global scope. Also, each class
represents types using a separate enum.)
Change-Id: Idef2141d3a4860529719382840a157b7b0c41643
Bug: skia:13457
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553595
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
The tests themselves are the same, just split up into their logical
groupings. http://go/unit-testing-overview#properties
"Focused. Above all, unit tests are narrow in scope, validating the
correctness of individual pieces of code rather than the correctness of
the system as a whole."
Change-Id: I6149536e84763abc98dfe243d4090bb25a555525
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553592
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Graphite's uniforms (SkUniform) distinguish between "Float" and "Half"
as separate types, so we now preserve this information in a Uniform
flag. We didn't have any tests verifying the behavior of Uniform flags
in SkRuntimeEffect at all, so I added a test here.
Change-Id: If6323ce26eeca2ef7da6c7424ad7350c6a0b7362
Bug: skia:13405
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553594
Reviewed-by: Brian Salomon <bsalomon@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Change-Id: I7abba20991482d72ca90b52a92034b9875be4098
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553578
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
This reverts commit 64284e1482.
Reason for revert: unused code
Original change's description:
> experimental alternative isconvex
>
> Bug: skia:
> Change-Id: I55175a95d37aad9a656cd211fc6c7238bdb7d674
> Reviewed-on: https://skia-review.googlesource.com/c/173361
> Commit-Queue: Mike Reed <reed@google.com>
> Reviewed-by: Mike Reed <reed@google.com>
Change-Id: Ib00acfc0964ecfd2fe01689c0d46d0773d5c5258
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/553588
Commit-Queue: John Stiles <johnstiles@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
